Chlorination of Acetaldehyde

chloroethane) has provided the main reason for an increased interest in the production of chloral ( , , -trichloroacetalde- hyde). The long-establishe...
0 downloads 0 Views 477KB Size
ENGINEERING AND PROCESS DEVELOPMENT

Chlorination of Acetaldehyde WILLIAM T. CAVE1 Reseorch Laboratories, Shawinigan Chemicals ltd., Shawinigan FaL, P . Q., Canada

T

HE demand for the insecticide DDT (dichlorodiphenyltri-

e

.

chloroethane) has provided the main reason for a n increased interest in the production of chloral (a,a,a-trichloroacetaldehyde). The long-established method of manufacturing chloral consists of the chlorination of ethyl alcohol, and because this compound has so many other vital uses, t h e search for a n alternative raw material is of practical concern. For reasons of both economy and chemical constitution acetaldehyde was selected for investigation. In 1832 Wohler and Liebig (7) had demonstrated the production of acid chlorides through the chlorination of aromatic aldehydes. Wurtz (8) confirmed these data b y his trials i n 1857 with aliphatic aldehydes, when he introduced the aldehyde into large flasks filled with chlorine gas in direct sunlight. The object of his experiments, however, h a d been the preparation of chlorala,a,a-trichloroacetaldehyde. Further trials in 1872 (9) were directed toward blocking the reactive -CHO group, so that chlorination could proceed at the methyl group. To obtain the reaction as he conceived it, alcohol, water, and hydrochloric acid were added to the aldehyde prior t o chlorination. Twenty to 30% conversion of aldehyde t o a mixture of di- and trichloroaldehydes indicated the partial success of his methods. Concurrently, in 1875, Pinner (6) was also concerned with the same problem. His procedures, which differed from those of Wurtz, involved the use of aldehyde or paraldehyde and low initial temperature for chlorination. The removal of hydrochloric acid was felt imperative t o prevent aldolization and the production of “croton chloral,” now called butyl chloral and known to be a,a,,B-trichlorobutyraldehyde. I n spite of claims t o a “main product,” the conversion of aldehyde to chloral could have been only 15y0by the best calculations from his data. The work herein described was essentially directed t o the chlorination of acetaldehyde a t the a-carbon position. Experimental Work

Preliminary trials using apparatus and methods described by Wurtz and Pinner confirmed the results as given above. Analyses for the reaction products were developed along the following lines. Compound

Basis of Test

CClsCHO CHsCHClCC12CHO

Caustic breakdown to chloroform Caustic breakdown to 1,l-dichloropropene Caustic breakdown and estimation of chloride Titration of total acids and estimation of HC1 as chloride Monoethanolamine breakdown and estimation of chloride by Fsjan’s method Karl Fischer method

CHC12CHO CHaCOOH CHCleCOOH He0

1 Present address, Central Research Department, Monssnto Chemical Co., Dayton 7, Ohio.

September 1953

The data given in the tables are for conversion of aldehyde, as these were the values of practical concern. The weight percentage figures obtained by analysis (the sums of which were found = 100 f 1.50/0)were converted into corresponding aldehyde weights and then to per cent conversion. This method accounts for t h e fact t h a t the total for each table is exactly 1 0 0 ~ o . The course of this investigation and its results were dependent essentially upon modifications of apparatus and procedure, t h e chemical reactants remaining little changed. Results are therefore grouped in terms of the equipment and method of its use, the table and figure captions indicating the systems involved. Using apparatus such as the first stage shown in Figure 3, two major changes were made i n technique: (1) Dispersion of chlorine was greatly improved and (2) the rate of introduction of the gas was increased. I n general, the interdependent rates of addition of chlorine and increase of kettle temperature were maintained at as high a level as was consistent with just providing for a n excess of chlorine in the solution. Specifically, in the initial portion of the run, where absorption appeared instantaneous, the chlorine was introduced as fast as the heat of reaction could be removed. Experience imposed a n upper limit of 10’ C. for this portion of the reaction. The accuracy of control was 322’ C. The most successful of these runs was one in which 310 grams of acetaldehyde mixed with 200 grams of water were chlorinated (excess chlorine at all stages) a t temperatures from 4’ to 92” C. in 43 hours; 675 grams of product boiling a t 95-97’ C. wereobtained, which represented a 65% conversion of aldehyde to chloral. The main by-product was butyl chloral. Record of chlorine absorbed, temperature, and time are given in Figure 1. Examination of per cent conversion results from a series of such runs indicated t h a t between the limits I to 3 moles of water per mole of aldehyde the reaction was not sensitive to the initial amount of water. It was decided to fix on the ratio 3 t o I wateracetaldehydebecausepurification distillations were not complicated with tarry residue, as they tended to be when the ratio was lower. T o improve further the efficiency of chlorine dispersion, the apparatus shown i n Figure 2 was built. The novelty consisted of introducing the gas in between two centrifugal pumps connected i n series. The analysis of a typical run is given in Table I. Subsequently, the water in the starting mix of 3 moles of water and 1 mole of aldehyde was replaced by 36y0 hydrochloric acid (Table I). The apparatus shown in Figure 3 was used to determine t h e results of carrying out the chlorination on a continuous basis. A 3 t o 1 water-acetaldehyde mix was fed continuouslyinto the first flask at about 0’ C., and the final product was collected from t h e fourth flask, which was kept at about 90’ C. A typical result is given i n Table 11. Before the obvious difficulties of continuous third-stage chlorination were considered, a series of runs was made, chlorinating to dichloroaldehyde continuously, using the two-pump dispersion system and finishing the third stage batchwise. I n this appara-

INDUSTRIAL AND ENGINEERING CHEMISTRY

1853

ENGINEERING AND PROCESS DEVELOPMENT Table 1.

each representing 20 liters of crude product. Conversion of acetaldehyde t o chIora1 ranged from 84 to 91yo: the limiting analyses for the runs are shown in Table 111. To carry out the third stage continuously, modifications mere made in the apparatus to adjust the holdup time. A determination of reaction rate constants gave K = 0.009 min.-1 down to 3% dichloroaldehyde; thereafter, K = 0.003 min.-l Experimental data were also considered i n terms of equations given by XacMullin and Weber ( 5 ) ,who dealt theoretically with the degree of completion to be expected from a first-order reaction carried out continuously in one kettle. The agreement waq very good and using these data the optimum third-stage holdup time was determined to be 32 hours.

Large Scale, Single-Stage, Batchwise Runs Conversion Aldehyde,

to

%

Chloral Butyl chloral Dichloroaldehyde Acetic acid Dichloroacetic acid

Table

II.

62.6 1 4 5.6 28.7 1.7

Small Scale, Four-Stage, Continuous Runs Conversion to Aldehyde,

% Chloral Butyl chloral Dichloroaldehyde Acetic acid Dichloroacetic arid

Table 111.

/

VENT

GAS SEPARATING BAFFLE

63.0 0 20.9 5.0

11.1

Large Scale, Three-Stage, Semicontinuous Runs Conversion_of Aldehyde,

ING CENTRIFUGAL

, Table TV.

Large Scale, Third-Stage, Batchwise and Continuous Runs

2.9

3.0

2.6 2.9

2.1

3.5

CIRCULATING

CENiRiFUGAL

Figure 2. Apparatus for Large Scale, Single-Stage, Batchwise Runs

Conversion of Aldehyde, % Batch Continnous 92.0 91.0

Chloral Butyl chloral Dichloroaldehyde Organic acids as acetic

!,’-

Using samples of stock crude dichloroacetaldehyde made on the continuous basis, the third stage was now completed continuously and batchwiee for comparison (Table IV).

ius, stage 2 consisted of a 5-liter flask fitted with a sintered-glass diffuser for the chlorine gas. The inlet for the overflow from stage 1 was a t the bottom of the flask near the diffuser and the overflow was near the top of the flask. A series of six runs was made,

Discussion

From the trials with the small scale single-stage, batrhwise apparatus it was clear that water acted effectively to block the reactive -CHO group to permit chlorination onlv at the a-carbon atom. Moreover, the data in Figure 1 indicated t h a t the chlorine substitution occurred in three distinct steps characterized by specific temperature limits. The reaction, therefore, was as follows: OH FIGURE 1.

CHaCHO

/ + H20 +CHaC-€I \

OH

OH

/

CH8CH

BATCHWISE

‘OH CHLORINATION OF ACETALDEHYDE

0

5

10

I5

20 TIME

1854

25 IN HOURS

M

35

OH

/ +- 3‘212 ---+ CCIaCH + 3HC1 \

OH

Chloral hydrate

45

The analyses indicated the main by-product of the reaction was butyl chloral, and it was clear t h a t t o improve further the conversion of aldehyde t o chloral the nature of this competitive reaction would have to

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 45, No. 9

ENGINEERING AND PROCESS DEVELOPMENT be understood. The pronounced beneficial effect of increasing the rate of input of chlorine at 0' to 10" C. suggested t h a t the side reaction must involve unchlorinated aldehyde and monochloroaldehyde. The probable reaction was visualized as follows:

OH

OH

CHSCHO

/

/

\

\

+ CHaClCH --+ C I L C H ( 0 H ) C H C l ~ H OH

OH

-

4+

CH8CI-I: = CCI H

O 'H

H2O

H

O 'H

OH

/ CHaOH = CClCH \

H&

OH

OH

/ [HCI]

CHsCH(0H)CHCICH

chlorine had t o be carefully controlled t o prevent substantial osidation of the aldehyde in all three stages, even though hydrochloric acid was present in saturation amounts. The high dichloroaldehyde figure indicated t h a t the holdup time for stage 3 was insufficient. Working on the problem of reducing the amount of dichloroacetic acid, trials were made with the larger apparatus with the two-pump system.

OH

/

+ Ch ---+ CHsCHClCCLCH \H

OH

STAGE I.

Butyl chloral hydrate Favorable conditions were therefore defined as those which increased the probability of aldehyde reacting with chlorine and decreased the probability of aldehyde reacting with monochloroal(:ehyde. It was with this conclusion in mind t h a t the two-pump dispersion system (&own in Figure 2) was built. The results in Table I1 were therefore anticipated regarding butyl chloral. This by-product had been effectively avoided. However, the very large increase in acetic acid was a new and undesirable feature. It seemed most likely that the oxidation was accomplished through the agency of a too large excess of chlorine: Clz

+ HzO HOCl

&

HOCl HCI

Apparatus for Small Scale, Four-Stage, Continuous Runs C.

S. 1.

Condenser Mercury seal stirrer Thermometer

+ HCI

+ (0)

Assuming this explanation t o be true, the water in the starting mix was replaced with 36% hydrochloric acid t o suppress the for-

mation of hypochlorous acid. The data of Table I indicated t h a t the production of acetic acid was significantly reduced, but a n increase in the formation of butyl chloral was also evident. Therefore, although the presence of hydrochloric acid appeared t o have s u p p d hypochlorous acid formation, it also served t o enhance the condensation reaction leading t o butyl chloral. h'onetheless, using a low initial temperature (-0" C.) and an excess of chlorine, and having hydrochloric acid present in the starting mix, a n 80% conversion of aldehyde to chloral was shown to be possible batchwise. T o examine what further possibilities might be inherent with a continuous process, the apparatus of Figure 3 wm constructed. From the data of Table I1 it is secn t h a t butyl chloral waa reduced t o less than 1%. This was the result expected, as the concurrent introduction of aldehyde and chlorine offered the best conditions for the prevention of the condensation reaction. The acetic acid figure was no higher than the batchwise trials using 36% hydrochloric acid t o start with, so that stage 1 at 0" t o 10' C. appeared t o have operated very satisfactorily. The dichloroaldehyde and dichloroacetic acid figures, however, indicated t h a t the other two stages a t 20' t o 30" and 70" t o 90" C. were not properly managed. Going back to the explanation for the presence of acetic acid, i t was expected t h a t the dichloroacetic acid indicated too great an excess of chlorine in stage 2 and possibly 3. T o find out whether this explanation was tenable, a solution of 75% chloral and 25% water was held at 85" C. for 13 hours, while chlorine was passed through the solution through a sintered-glass diffuser. Thirteen per cent trichloroacetic acid was separated (from the high boiling residue) as a pink-white solid melting at 50-55" C. Thus i t appeared that the excess of September 1953

Figure 3.

As can be seen from Table 111, the objective of the trials was successfully achieved with the best conversion figures of 90% for chloral and less than 3% for all other by-products, with the exception of dichloroaldehyde which ran a t 7 %. Having discovered a means of controlling the condensation and oxidation side reactions, a final attempt was made t o determine whether a completely continuous operation could be used. Bell and Longuet-Higgins ( 1 ) have considered the halogenation of acetone, using alkaline hypobromite, and found that the reaction was of zero order respecting the hypobromite and of first order respecting acetone. I n this investigation, as the rate of chlorination is independent of the concentration of chlorine, it is assumed t o be of first order respecting the concentration of dichloroaldehyde. On this basis the reaction rate constants were determined experimentally t o be K = 0.009 min.-l down to 2.7% by weight of dichloroaldehyde and K = 0.003 min.-I below this amount. This result was confirmed by substituting the experimentally determined degree of completion of the reaction in stage 3 in the equations of MacMullin and Weber. A reaction velocity constant of 0,009 rnin.-l was the answer found. ' Table IV shows that when the holdup time waa properly adjusted, results were entirely comparable with batohwise trials. The lowest practical limit of dichloroaldehyde was ahown to be near 3% on the basis of aldehyde conversion. Conclusions

I n the chlorination of acetaldehyde or paraldehyde in the presence of water it is possible t o exceed 90% conversion t o a,adichloroaldehyde or a,a,a-trichloroaldehyde. The main competing reaction appears to involve the condensation in the presence of hydrochloric acid of a molecule of acetaldehyde with a molecule of monochloroaldehyde and produces finally a,a,p-trichlorobutyraldehyde. Another side reaction involves the o x i d a

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

1855

ENGINEERING AND PROCESS DEVELOPMENT tion, through the agency of hypochlorous acid. of the aldehydes t o the corresponding acetic acid, dichIoroacetic acid, and trichloroacetic acid. Both types of aide reaction can be controlled through proper arrangement of temperature and conditions pertaining t o the introduction of chlorine. The chlorination can be carried batchwise, but is most efficiently managed on a semicontinuous or completely continuous basis. Acknowledgment

Grateful acknowledgment is made t o Shawinigan Chemicals, Ltd., for permission t o publish; t o K. G. Blaikie for the helpful direction under which this work was carried out; t o T. P. Shaw for providing the analytical data; to D. J. Kennedy and G. E. Haddeland ( 3 , 4 )for the development of the tu-o-pump dispersion system; t o J. H. Alexander ( 9 ) for the investigation of holdup times for the third stage of the continuous system; and finally t o Armand Xadeau for able technical assistance.

literature Cited (1) Bell, R. P., and Longuet-Higgins, H. C., J. Cfiem. S o c , 1946, 636. (2) Cave, W. T., Alexander, J. H., and hlcCoubrey, J. A. (to Shawinigan Chemicals, L a . ) , .U. S. Patent 2,606,864 (Aug. 12, 1952); Brit. Patent 634,938 (March 29, 1950). (3) Cave, R. T., and Haddeland, G. E. (to Shawinigan Chemicals, Ltd.), U. 9. Patent 2,552,934 (May 15, 1951): Brit. Patent 644,914 (Oct. 18, 1950). (4) Haddeland, G. E., and Kennedy, D. J. (to Shawinigan Chemicals. Ltd.). U. S. Patent 2,521,215 (Sept. 5. 1950): Can. Patent 470,578 (Jan. 2, 1981); Brit. Pakkt 633,378 (Dec. 12, 1949). (5) AlacMullin, R. B., and Weber, &I., Tram. Am. Inst. C h e n . Engrs., 31, 409 (1935). (6) Pinner, A., Liebigs Ann. Chenz., 179, 21 (1875). (7) Wohler, F., and Liebig, J., Ibid., 3, 262 (1832). (8) Wiirtz, A., Ibid., 102, 93 (1857). (9) Wiirtz, A., and Vogt, G., Compt. vend., 74, 777 (1872). RECEIVED for review January 21, 1963. A C C E P T E D hfay 4, 1963. Presented at the annual aonvention of the Chemical Institute of Canada, Halifax, N. S., June 1949.

Kinetics of Coal Gasification Proposed Mechanism of Gasification HOWARD R. BATCHELDER' AND ROBERT M. BUSCHE2, U. S. WILLARD P. ARMSTRONG, Wcishingfon University, St. Louis,

AND

I

N THE course of Fork in the development of the synthetic

liquid fuels program, attention has been centered primarily on coal as a source material for the necessary synthesis gas or hydrogen. Gasification of coal has been undertaken in a number of wags; most of the work of the Bureau of R h e s has been concentrated on the reaction in dilute suspension of powdered coal with oxygen and steam. Over a period of years, considerable scientific labor has been devoted to small-scale studies of certain single reactions t h a t are a part of the complex reaction system comprising coal gasification. Isolation of these reactions was, of course, necessary to obtain adequate and accurate data on reaction mechanism and rate, although in many cases competing reactions and the inherent difficulties of high temperature measurement required very ingenious experimental methods. Although such studieE serve as a basis for the understanding of the nature of the reactions involved, they do not integrate the kinetics of the individual reactions with the heat transfer and mass transport phenomenon, which are necessarily a part of the gasification process. Because of the significant influence of size on these phenomena, bench-scale x o r k has not been adequate for application to commercial-size units. On the other hand, empirical treatment of data obtained from the observation of pilot plant scale gasifiers has resulted in considerable information regarding the 1

2

Present address, Battelle Memoiial Institute, Columbus, Ohio. Present address, E. I. du Font de iiemours & Co., Ino., Belle, W. Va

1856

Bureau o f Mines, Louisiana, M o . , Ma.

operation and performance of each individual unit but does not clarify the complex physical processes involved and does not permit the design of other gasifiers in other than empirical fashion. For these reasons, it was decided to make this work an engineering study of the gasification process as a whole. An extensive search of the literature was made to obtain available rate data and to suggest a suitable mechanism of the over-all process, including heat and mass transfer. Theoretical equations expressing the kinetic changes in gasification were derived and applied in the calculation of conversion and product distribution to be expected in the gasification of powdered coal with such variable operating conditions as steam-coal ratio, oxygen-coal ratio, gasifier heat loss, and steam preheat. The results of this theoretical approach were then compared with data obtained experimentally from operation of suitable gasifiers. The theoretical calculations depend on the results of the literature survey, but not to the extent that might be thought. Three specific rate constants are involved, but the accuracy of these constants does not affect the form of the equations derived, and consequently does not affect the correlation of the chemical reactions and thermal and mass transfer involved in gasification. -4n extensive bibliography of the gasification literature has been published (44). The thermodynamics of the gasification system have been fairly well defined and may serve as a guide in choosing the variables to be studied and t o indicate the limits toward which a reacting system mag progress. Knmerous papers have presented

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 45, No. 9